Business Technology: Will Windows 2003 Play Well With Others?
Just over six years ago, on April 7, 1997, InformationWeek ran a cover story with a photo of Bill Gates and the headline, "The Billion He Really Wants." Here's what the teaser to that story said: "A billion NT database transactions a day, that is. At the same time, Gates wants to support 10,000 concurrent users and a trillion bytes of data. But so far, SQL Server handles fewer than 12 million transactions a day--just 1% of the billion-a-day target. Can Microsoft close the gap while ensuring stability?" (The story is still online: informationweek.com/625/25iumic.htm.)
At the time, Gates' projection drew more than a little ridicule from competitors, and it also generated skepticism from some objective observers whose doubts were centered not so much on Microsoft's capabilities but rather on the sheer physical boundaries that would have to be stretched: Can microprocessors scale to that level? If the software designed for those chips can reach those speeds, will it be stable? Or, if the software designed for those chips can be fully stable, can it also reach the required speeds? If both objectives can be achieved, will the resulting products be competitive on cost with larger systems that were already achieving such performance? Is it possible that they could be less expensive? And if so, could they be significantly less expensive and thereby help justify a customer's decision to switch to a new and relatively untested architecture?
According to a dozen or so military men I spoke to, Rumsfeld simply failed to anticipate the consequences of protracted warfare. He put Army and Marine units in the field with few reserves and an insufficient number of tanks and other armored vehicles. Pentagon officers spoke contemptuously of the Administration's optimistic press briefings. "It's a stalemate now," the former intelligence official told me.
-- Seymour Hersh, The New Yorker, April 7
Well, as it has proven throughout its history, Microsoft is nothing if not persistent. In a story on our Web site last Thursday about the company's release of Windows Server 2003 (informationweek.com/937/win2003.htm), InformationWeek senior writer Aaron Ricadela noted that with the same-day introduction of a 64-bit version of its SQL Server 2000 Enterprise Edition database, "Microsoft hopes to put to rest questions about the scalability of its software that have dogged the company for years." Indeed, Ricadela reported, last Thursday also saw a Windows system, for the first time ever, take the No. 1 spot on a respected third-party ranking of the fastest data-processing machines in the world. "According to the Transaction Processing Council's single-system benchmark for database transaction systems, a Hewlett-Packard Superdome server running Windows Server 2003 and SQL Server--and with 64 of Intel's upcoming 'Madison' Itanium 2 chips inside--performs 658,277 transactions per minute, a new record," Ricadela reported.
That's about 39.5 million transactions per hour, which would equate to 948 million per day--not quite the billion forecast six years ago, but only 5% short. And for businesses intrigued by such possibilities, there's more news: Intel president Paul Otellini says that within the next two years, enhanced Itanium chips will double or triple that performance.
All quite impressive, but beyond the technology, some real-world questions remain. For one, what about Linux: Will Microsoft make it easy for customers to blend its dazzling new Windows systems with the Linux-based servers rapidly gaining acceptance in large organizations? Or will the company inject its feisty competitiveness into its customers' daily lives by making such compatibility painful and complex? Will it adapt its unpopular licensing structures to conform to the strapped budgets of its customers and prospects? Will Microsoft complement these powerful new building blocks with integration tools and strategies that offer more options than simply rip and replace? And while its unfolding suite of .Net technologies will no doubt be optimized to work with Windows Server 2003, will it also work smoothly with non-Windows systems?
Microsoft needs to make sure it supplies its customers with the right answers to those questions. Because if it doesn't, the past six years' significant technical advances will amount to little more than a Pyrrhic victory.
Building A Mobile Business MindsetAmong 688 respondents, 46% have deployed mobile apps, with an additional 24% planning to in the next year. Soon all apps will look like mobile apps – and it's past time for those with no plans to get cracking.